Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 55
GOALS Global Ocean-Atmosphere-Land System for Predicting Seasonal-to-Interannual Climate: A Program of Observation, Modeling, and Analysis 5 Data Management The GOALS program would require a robust data management system. The data management plan would likely develop from current TOGA and post-TOGA data management plans. Being a broader study, however, GOALS would encompass new data sets from field experiments, specialized data sets, and data sets derived from new observational networks designed to complement existing global measurements. The amount of data used and produced by GOALS is expected to be significantly greater than that used and produced by TOGA. Moreover, unique requirements for GOALS would probably arise with the introduction of new operational observation platforms or techniques, portable self-describing data formats, novel data transmission technologies, and the development of technology such as larger mass storage systems. The benefits of cross-program science transfer and the likelihood of tight budgets in the future make paramount the need to share resources with other programs. Using facilities and expertise that are already developed and proven ensures both economy of effort and continuity. Also, the transition from a research to an operational mode that is proposed for GOALS has impacts on the budget process, funding sources, and data management responsibilities. PRINCIPLES AND OBJECTIVES Data management requirements should be driven by the science and operational requirements for data streams for both near-real-time and
OCR for page 56
GOALS Global Ocean-Atmosphere-Land System for Predicting Seasonal-to-Interannual Climate: A Program of Observation, Modeling, and Analysis retrospective use. Data streams and data sets would be variable, with the type and format of the data differing, depending on their origin (such as new and existing observing systems, field research programs, and experimental programs). Attention to data quality and continuity would be imperative. During the 15-year time frame of the program, hundreds of gigabytes of data would be collected and analyzed—and then used well beyond the lifetime of the program. (Data sets from the GARP Atlantic Tropical Experiment, for example, are still being used many years after the termination of the experiment.) Clearly, it would be essential that data sets and data streams from GOALS be safely archived and remain easily accessible to the research community. As the number of users of data sets evolved—including modeling and satellite data sets, and in situ measurements such as those from TOGA-COARE and the Global Climate Observing System—the GOALS data management system would need to be able to handle new data streams. A data management plan to accommodate near-term efforts and to provide a framework for long-term goals would then be required. This would be particularly important with the emergence of the synergy between GOALS and other international programs that address global dynamics (for example, GEWEX, GCOS, GOOS, ACCP, WOCE, JGOFS). The following list of principles and objectives should guide the development of a detailed GOALS data management plan: Data are distributed as soon as possible, and there is full and open exchange of data and derived products. Reanalysis is an integral part of the program; this necessitates assessable retention of observations in original form. Sufficient metadata (quality control flags, instrument history, processing attributes, and so on), are provided so that researchers can evaluate the usefulness of the data. An ongoing, collaborative effort is maintained between data managers and researchers to assess standards for the production of data sets. Some data and data display tools are available on-line and in near real-time. Backup data sets are archived to accommodate any change in parameters or algorithms that would affect derived quantities. Other data management activities, such as the EOS Data and Information System (EOSDIS) are monitored and evaluated for their relevance and applicability to GOALS.
OCR for page 57
GOALS Global Ocean-Atmosphere-Land System for Predicting Seasonal-to-Interannual Climate: A Program of Observation, Modeling, and Analysis Real-time and delayed-mode four-dimensional data-assimilation systems provide fields to complement observations. There is a distributed data archive and access system involving the full range of regional, national, and world data centers. An on-line service is created with high-level project information (for example, the progress of an experiment, program information, availability of data), perhaps similar to other information systems currently in use, such as those for WOCE, TOGA-COARE, and Storm-scale Operational and Research Meteorology (STORM). DATA CENTERS Considerable infrastructure already exists to process, archive, and distribute Level I (radiance and similar unprocessed measurements), Level II (geophysical quantities), and Level III (assimilated or blended analyses) data sets. However, that infrastructure must be maintained and, in many instances, expanded. Similar to TOGA's data centers, existing institutions would be relied upon to a large extent to manage specialized data streams and data sets. Examples already exist in the TOGA program, where existing institutions have taken on the task of quality control and general support, such as: The Tropical Sea Level Data Center (Hawaii) produces Level II-B sea level data, The Tropical Ocean Subsurface Data Center (IFREMER, Brest, France) generates Level II-B subsurface temperature and salinity data, The European Centre for Medium-Range Weather Forecasts (ECMWF) processes atmospheric Level III-A global meteorological fields, The TOGA Marine Climatology Data Center (U.K. Meteorological Office) produces Level II-B marine surface meteorological data, The Global Sea-Surface Temperature Data Center (NMC) performs Level II-A, III-A, and III-B processing of SST observations, and The TOGA Upper-Air Data Center (New Delhi) has responsibility for upper-air observations. Complementing these data centers are several TOGA-related institutes, universities, and national services that hold or produce data
OCR for page 58
GOALS Global Ocean-Atmosphere-Land System for Predicting Seasonal-to-Interannual Climate: A Program of Observation, Modeling, and Analysis sets of interest to TOGA researchers. Examples are NOAA's Atlantic Oceanographic and Meteorology Laboratory (AOML) for surface current drifter data; NOAA's Pacific Marine Environmental Laboratory (PMEL) for near-real-time data from the TOGA-TAO array; Florida State University for pseudo-wind stress fields; and WOCE Data Assembly Centers for various types of mooring, float, and drifter data, and sea-level and upper-ocean thermal data. In most cases, World Data Centers A and B for meteorology and oceanography are the final (delayed-mode) repository for all TOGA data sets. To investigate the feasibility of predicting short-term climate variations, quality-controlled meteorological and oceanographic data from established data centers would be required to initialize forecast models. This would include the analysis of multivariate fields and considerable data processing and production activities, that is, data assimilation and modeling. Model initialization would require the merging and gridding of various fields through multivariate optimal interpolation or other techniques. This operation would entail an evolving set of quality-control procedures. Model runs would create large quantities of output that would have to be documented, archived, and distributed on various media. To achieve these goals, advanced data archiving, retrieval, and visualization capabilities would be required for both assimilation and forecast products. The dissemination of these data and information throughout the research community, and to other pertinent interests, would be of paramount importance to the program. This dissemination capability would require investments for leveraging existing and planned interoperable data systems that could provide GOALS scientists with access to data sets via common interfaces. The communication backbone for such a system must be capable of transmitting millions of bits per second. To achieve its objectives, GOALS is very dependent on ongoing efforts within NASA, NOAA, National Science Foundation, U.S. Department of Defense, and U.S. Department of Energy data centers for the necessary infrastructure to support such a data management system. It is hoped that this in place network of data processing centers and data repositories would be able to provide expeditiously the GOALS program with some important state-of-the-art data management capabilities. A major commitment of the GOALS program would be the assembly and analysis of a 30-year (1971-2000) climatic data set to examine seasonal-to-interannual variability. This effort would involve developing ways to incorporate remotely sensed data, as well as multivariate, relatively and absolutely inhomogeneous, and sparse data
OCR for page 59
GOALS Global Ocean-Atmosphere-Land System for Predicting Seasonal-to-Interannual Climate: A Program of Observation, Modeling, and Analysis into coupled ocean–atmosphere–land models. The resulting data sets would also facilitate empirical studies of the predictability of the coupled system. Strong communication links and interoperability between major data centers and data-distribution centers could be required.
OCR for page 60
GOALS Global Ocean-Atmosphere-Land System for Predicting Seasonal-to-Interannual Climate: A Program of Observation, Modeling, and Analysis This page in the original is blank.
Representative terms from entire chapter: